L4S-Aware Network Coordination

4 4 4 4 Low Latency Low Loss Scalable Throughput (LS) is a technology intended to reduce queue delay problems, ensuring low latency to Internet Protocol flows with a high throughput performance. As of now, widespread implementation LS is not common and only some equipment supports LS. Thus, the devices that do not support LS will likely have degraded user experience during congestion, such as packet losses, and add to the congestion of the overall network.

4 4 4 4 4 Exemplary embodiments described herein include systems, methods, computer readable medium and processing nodes for LS aware network coordination and congestion management. An exemplary method includes receiving, by a wireless network communicatively connected to an LS capable access node and a non-LS capable access node, a congestion signal from an LS capable wireless device. The method further includes, in response to receiving the congestion signal from the LS capable wireless device, adjusting, by the wireless network, network configurations for a non-L4S capable access node.

In the following description, numerous details are set forth, such as flowcharts, schematics, and system configurations. It will be readily apparent to one skilled in the art that these specific details are merely exemplary and not intended to limit the scope of this application.

There are a wide range of applications that rely on real-time communication, interactive experiences, or high-performance data transmission such as online and cloud gaming, video conferencing, AR/VR, live streaming, and others. All these applications can potentially benefit from the use of L4S technology.

4 LS utilizes ECN (Explicit Congestion Notification) embedded within the IP header of an IP packet to signal queue congestion within a radio access network (RAN) to an application. This congestion information is handled by scalable congestion control algorithms at both the sender and receiver ends and communicated to the application server, prompting adjustments to the application's bitrate. As a result, it aligns with the capacity of the established communication link.

4 4 4 4 4 4 As of now, widespread support of LS is not common and only some latest generation of access nodes (base stations) support LS. It may take a significant time and effort to upgrade hardware on the base stations to fully support LS across all carriers. LS information received by LS supported access nodes can be leveraged for network-wide coordination with non-supporting base stations to enhance overall congestion management, even if some access nodes do not support LS currently.

5 4 4 4 4 The Network Data Analytics Function (NWDAF) in the core, such as theG core, can aggregate LS-based congestion metrics (such as latency, ECN marking levels, and packet loss) from LS capable access nodes and create a global view of network congestion patterns. This data can then be communicated to both LS and non-LS capable access nodes.

4 4 NWDAF can apply predictive analytics to anticipate congestion propagation and provide recommended actions to non-LS access nodes, such as reducing uplink (UL) configurations, rebalancing traffic, or shifting non-LS capable devices to lower-bitrate slices in congestion-prone areas.

4 4 4 4 Centralized self-organizing network (SON) functions coordinate congestion-related adjustments across both LS and non-LS capable access nodes. By receiving congestion alerts from LS-enabled cells, SON can adjust cell boundaries, optimize handover thresholds, and direct traffic offload to underutilized non-LS supported cells.

4 4 4 For example, SON facilitates cell boundary adjustments in non-LS supported cells based on anticipated congestion in LS cells, helping to balance network load and reduce congestion impacts on non-LS supported wireless devices, especially in dense urban or high-traffic environments where maintaining stable performance is critical.

4 4 4 4 4 Intent-based radio access network (RAN) automation extends LS-driven congestion responses across non-LS supported cells. By interpreting high-level intents, such as “minimize latency impacts during high congestion,” the network can adapt slice configurations, prioritize essential traffic, and reduce UL layers where possible, even if non-LS supported cells cannot process ECN signals directly. This allows non-LS cells to benefit from congestion patterns observed in LS-supporting cells, dynamically adjusting resources for continuity and quality of service (QoS) consistency.

4 4 4 4 An artificial intelligence/machine learning (AI/ML) model analyzes congestion patterns from LS capable access nodes and generate predictions or configurations that apply to non-LS supported cells. For example, historical LS congestion patterns are used to forecast expected loads across nearby non-LS supported cells.

4 Historical congestion data from LS capable devices, including metrics like packet loss, delay, and throughput, can be aggregated along with network conditions (e.g., time of day, traffic loads). Key features might include frequency and duration of congestion events, traffic volume during peak hours, latency or packet loss trends by location or slice, and the like.

4 Accordingly, such ML models are used to suggest configurations for non-LS capable access nodes, such as prioritizing low-bitrate slices or adjusting power control settings in UL. These adjustments are made dynamically via the backhaul to distribute load and mitigate congestion impact.

4 4 4 4 Policy-based control in the core uses LS congestion data to optimize resource allocation across LS and non-LS supported cells. The core adjusts resources for non-LS access nodes by lowering priority on non-essential traffic during congestion spikes or adjusting UL and download (DL) resource allocation to optimize for stability.

4 4 4 4 4 4 4 4 4 Policies prioritize handovers of non-LS capable devices to non-LS capable cells when LS-enabled access nodes detect congestion, allowing the core to respond effectively even if non-LS capable cells lack direct LS functionality. By offloading non-LS capable wireless devices to neighboring non-LS capable cells, resources on the LS capable cells are freed, allowing LS-capable wireless devices to benefit more from the reduced congestion and efficiently manage their bitrate in response to ECN signals.

4 4 4 There are also access nodes that do not support LS that would benefit from receiving congestion information from nearby LS supported wireless devices. A cooperative network environment is provided where LS supported wireless devices share network status and congestion information to improve user experience and network performance.

1 5 FIGS.- These and other examples will be described in greater detail below in relation to.

1 FIG. 100 100 101 102 170 4 120 4 122 depicts an exemplary systemfor network congestion management. Systemincludes a communication network, a core network, a radio access network (RAN)and LS capable wireless devicesand non-LS capable wireless devices.

102 101 111 102 5 103 5 103 5 5 103 102 5 103 Core networkis connected to communication networkover communication link. Core networkincludes a 5G core (GC).GCas used herein are core network components used for managing data forG networks. In embodiments,GCmay include an evolved packet core (EPC), used for managing data for LTE, 4G and /or other networks. In instances, the core networkmay have other types of core architecture (e.g., 6G core architecture) that at least perform some similar functions as and/or share at least some components with theGCwith respect to congestion management and network slicing for wireless devices.

102 It should be noted that core networkmay include other components used for managing data for networks not described herein, such as a satellite core network.

5 103 105 105 4 120 4 122 105 4 120 4 122 105 105 4 122 In embodiments,GCincludes an access and mobility function (AMF). AMFreceives connection and session related information from the LS capable wireless devicesand non-LS capable wireless devicesand is responsible for handling connection and mobility management tasks on a 5G network. In an embodiment, AMFmay be used for determining connection configuration for LS capable wireless devicesand non-LS capable wireless devices. For example, AMFmay communicate with a network slice selection function (NSSF) to determine a slice configuration based on changes in the quality of service (QoS) policies. In instances, AMFmay be used for assigning a network slice for non-LS capable wireless devices.

170 4 171 4 172 171 4 172 102 102 4 171 4 4 171 The RANincludes LS capable access nodesand non-LS capable access nodes. In embodiments, access nodesand non-LS capable access nodesinclude an evolved Node B (eNodeB) and a next generation Node B (gNodeB). As used herein, an eNode B is a base station in LTE/4G networks used for connecting a user device to core network. A gNodeB, as used herein, is a base station in 5G networks and/or other networks used for connecting a user device to core network. The gNodeB may include, for example, centralized units (CUs) and distributed units (DUs). In embodiments, the LS capable access nodesare equipped with LS-aware scheduling and dual-queue active queue management (DualQ AQM). For example, LS capable access nodesmay be configured to mark packets when congestion is detected in the network, such as through explicit congestion notification (ECN) marks.

170 102 112 170 170 120 102 RANis connected to core networkover communication link. RANmay include other devices and additional nodes not described herein. For example, RANmay include devices used for forwarding media files over IP from wireless devicesto core network.

100 120 100 120 122 121 120 122 120 122 120 170 113 113 Systemalso includes wireless devices. In embodiments, systemmay include two or more wireless devices. Wireless devicesandare configured to operate in one or more coverage areas. Wireless devicesandmay include an end-user wireless device. Wireless devicesandmay include any device configured to send and receive data. In embodiments, wireless devicecommunicates with RANover communication link. Examples of communication linkmay include 5G network, 4G LTE, and the like.

120 4 122 4 4 120 4 4 120 5 102 In embodiments, wireless devicesinclude at least one LS capable wireless device and wireless devicesinclude at least one non-LS capable wireless device. LS capable wireless devicesincludes devices configured to support LS traffic handling. In instances, LS capable wireless devicesare configured to transmit congestion signals toGC.

101 101 101 101 120 1 3 4 5 5 5 5 101 x Communication networkmay be wired and/or wireless communication network. In embodiments, communication networkmay include processing nodes, routers, gateways, physical and/or wireless data links for carrying data among various network elements, including combinations thereof. In embodiments, communication networkmay include a local area network, a wide area network, an inter-network, such as the internet, and the like. Communication networkmay be capable of carrying data, such as, for example, to support multimedia files, and data communications by wireless devices. Wireless network protocols can include multimedia broadcast multicast service (MBMS), code division multiple access (CDMA)RTT, Global System for Mobile communications (GSM), Universal Mobile Telecommunications System (UMTS), High-Speed Packet Access (HSPA), Evolution Data Optimized (EV-DO), EV-DO rev. A, Third Generation Partnership Project Long Term Evolution (GPP LTE), Worldwide Interoperability for Microwave Access (WiMAX), Fourth Generation broadband cellular (G, LTE Advanced, etc.), and Fifth Generation mobile networks or wireless systems (G,G New Radio (“G NR”), orG LTE), 6G and/or non-terrestrial networks. Wired network protocols that may be utilized by communication networkcomprise Ethernet, Fast Ethernet, Gigabit Ethernet, Local Talk (such as Carrier Sense Multiple Access with Collision Avoidance), Token Ring, Fiber Distributed Data Interface (FDDI), Asynchronous Transfer Mode (ATM), and/or so forth. Communication network 101 may also include additional base stations, controller nodes, telephony switches, internet routers, network gateways, computer systems, communication links, or some other type of communication equipment, and combinations thereof.

102 102 101 4 120 105 120 105 107 105 The core networkincludes core network functions and elements. The core networkmay be structured using a service-based architecture (SBA). The network functions and elements may be separated into user plane functions and control plane functions. In an SBA architecture, service-based interfaces may be utilized between control-plane functions, while user-plane functions connect over point-to-point link. The user plane function (UPF) accesses a data network, such as network, and performs operations such as packet routing and forwarding, packet inspection, policy enforcement for the user plane, quality of service (QoS) handling, etc. The UPF may detect congestion of a network, such as form a LS capable wireless device. In instances, the UPF may apply dual queue active queue management (DualQ AQM) based on the congestion detection. The control plane functions may include, for example, a network slice selection function (NSSF), a network exposure function (NEF), a network repository function (NRF), a policy control function (PCF), a unified data management (UDM) function, an application function (AF), an AMF, such as AMF, an authentication server function (AUSF), and a session management function (SMF). Additional or fewer control plane functions may also be included. The AMF receives connection and session related information from the wireless devicesand is responsible for handling connection and mobility management tasks. The SMF is primarily responsible for creating, updating, and removing sessions and managing session context. The UDM function provides services to other core functions, such as the AMF, SMF, and NEF. The UDM may function as a stateful message store, holding information in local memory. The NSSF can be used by AMFto assist with the selection of network slice instances that will serve a particular device. Further, the NEF provides a mechanism for securely exposing services and features of the core network.

120 105 105 105 105 In instances, the UDM may include a mapping of DNNs to network slice selection assistance information (nSSAI) associated with a wireless device. nSSAI includes a set of single nSSAI(S-nSSAI). Each S-nSSAI may include a slice/service type and a slice differentiator (SD). For example, AMFmay query UDM for S-nSSAIs associated with a DNN. In an example, AMFmay use NSSF for selecting a S-nSSAI based on additional requirements, such as regional availability. In some embodiments, UDM may detect a subscription expiration for a slice, such a validity period based slice, and notify AMFof the expiration. Once notified, AMFupdates nSSAI by removing expired S-nSSAIs.

102 102 102 5 103 Although one core networkis shown, multiple core networksmay be utilized. Alternatively, the single core networkmay include a distributed, cloud-native, converged core gateway. Thus, the converged core gateway could connect an EPC toGCnetwork.

111 112 111 112 1 1 5 6 111 112 111 112 Communication linksandcan use various communication media, such as air, space, metal, optical fiber, or some other signal propagation path, including combinations thereof. Communication linksandcan be wired or wireless and use various communication protocols such as Internet, Internet protocol (IP), local-area network (LAN), S, optical networking, hybrid fiber coax (HFC), telephony, T, or some other communication format - including combinations, improvements, or variations thereof. Wireless communication links can be a radio frequency, microwave, infrared, or other similar signal, and can use a suitable communication protocol, for example, Global System for Mobile telecommunications (GSM), Code Division Multiple Access (CDMA), Worldwide Interoperability for Microwave Access (WiMAX), Long Term Evolution (LTE),G NR,G or combinations thereof. Other wireless protocols can also be used. Communication linksandcan be direct links or might include various equipment, intermediate components, systems, and networks, such as a cell site router, etc. Communication linksandmay comprise many different signals sharing the same link.

170 4 171 4 172 170 102 120 170 102 120 170 102 120 In embodiments, RANmay include various access network systems and devices such as LS capable access nodesand non-LS capable access nodes. The RANis disposed between the core networkand the end-user wireless device. Components of the RANmay communicate directly with the core networkand others may communicate directly with the end user wireless device. The RANmay provide services from the core networkto the end-user wireless device. It is understood that the disclosed technology may also be applied to communication between an end-user wireless device and other network resources, such as relay nodes, controller nodes, antennas, etc. Further, multiple access nodes may be utilized. For example, some wireless devices may communicate with an eNodeB and others may communicate with a gNodeB.

171 172 171 172 171 172 172 In additional embodiments, access nodesandmay comprise two co-located cells, or antenna/transceiver combinations that are mounted on the same structure. Alternatively, access nodesandmay comprise a short range, low power, small-cell access node such as a microcell access node, a picocell access node, a femtocell access node, and/or a home eNodeB device. As will be further described below, functionality for network node switching may be included within the access nodesand.Access nodes 171 andcan be configured to deploy one or more different carriers, utilizing one or more RATs. For example, a gNodeB may support NR. It would be evident to one of ordinary skill in the art, in light of this disclosure, the many other combinations of access nodes and carriers that could be deployed.

171 172 The access nodesandmay include a processor and associated circuitry to execute or direct the execution of computer-readable instructions to perform operations such as those further described herein. Access nodes can retrieve and execute software from storage, which can include a disk drive, a flash drive, memory circuitry, or some other memory device, and which can be local or remotely accessible. The software comprises computer programs, firmware, or some other form of machine-readable instructions, and may include an operating system, utilities, drivers, network interfaces, applications, or some other type of software, including combinations thereof.

4 120 4 122 171 172 4 171 4 172 4 171 172 4 4 171 172 4 The LS capable wireless devicesand non-LS capable wireless devicesmay include any wireless device included in a wireless network. For example, the term “wireless device” may include a relay node, which may communicate with an access node. The term “wireless device” may also include an end-user wireless device, which may communicate with access nodesandthrough the relay node. The term “wireless device” may further include an LS capable wireless device that communicates with the access nodedirectly without being relayed by a relay node or a non-LS capable wireless device that communicates with access node. In embodiments, LS capable wireless devices may be able to communicate with access nodesorregardless of LS capability and non-LS capable wireless devices may be able to communicate with access nodesoragain regardless of non-LS capability.

120 122 171 172 120 122 120 122 Wireless devicesandmay be any device, system, combination of devices, or other such communication platform capable of communicating wirelessly with access nodesandusing one or more frequency bands and wireless carriers deployed therefrom. Each of wireless devicesand, may be, for example, a mobile phone, a wireless phone, a wireless modem, a personal digital assistant (PDA), a voice over internet protocol (VoIP) phone, a voice over packet (VOP) phone, or a soft phone, an internet of things (IoT) device, as well as other types of devices or systems that can send and receive audio or data. The wireless devicesandmay be or include high power wireless devices or standard power wireless devices. Other types of communication platforms are possible.

100 100 100 120 122 100 1 FIG. Systemmay further include many components not specifically shown inincluding processing nodes, controller nodes, routers, gateways, and physical and/or wireless data links for communicating signals among various network elements. Systemmay include one or more of a local area network, a wide area network, and an internetwork, such as the internet. Systemmay be capable of communicating signals and carrying data, for example, to support voice, push-to-talk, broadcast video, and data communications by end-user wireless devicesand. Systemmay include additional base stations, controller nodes, telephony switches, internet routers, network gateways, computer systems, communication links, other types of communication equipment, and combinations thereof.

100 170 102 Other network elements may be present in systemto facilitate communication but are omitted for clarity, such as base stations, base station controllers, mobile switching centers, and dispatch application processors. Furthermore, other network elements that are omitted for clarity may be present to facilitate communication, such as additional processing nodes, routers, gateways, and physical and/or wireless data links for carrying data among the various network elements, e.g. between the RANand the core network.

100 The methods, systems, devices, networks, access nodes, and equipment described herein may be implemented with, contain, or be executed by one or more computer systems and/or processing nodes. The methods described above may also be stored on a non-transitory computer readable medium. Many of the elements of systemmay be, comprise, or include computers systems and/or processing nodes, including access nodes, controller nodes, and gateway nodes described herein.

The operations for network congestion management may be implemented as computer-readable instructions or methods, and processing nodes on the network and/or computing device, such as end user wireless device, for executing the instructions or methods. The processing node may include a processor included in the access node or a processor included in any controller node in the wireless network that is coupled to the access node. The computing device may include at least a processor and a memory with instructions configuring the processor to execute instructions.

2 FIG. 200 200 202 202 170 102 101 202 4 202 Now referring to, an exemplary systemfor network congestion management is presented. Systemincludes wireless network. Wireless networkmay include a RAN, core network and/or a communication network, which may be the same as, respectively, RAN, core networkand communication network. Wireless networkincludes services and components used by a wireless network for LS Aware network coordination. In an example, wireless networkis configured to utilize scalable congestion control algorithms, such as TCP Prague.

200 220 220 120 220 4 222 220 4 222 4 222 4 4 222 4 222 4 4 4 222 202 202 4 222 4 222 Systemincludes wireless devices. Wireless devicemay be the same as wireless device. Wireless devicesinclude LS capable device. In embodiments, wireless devicesmay include a plurality of LS devices. LS devicemay include any LS capable device. For example, LS devicemay be any device capable of transmitting ECN-capable packets and receiving ECN feedback. In instances, LS devicemay use dual queue active management (DualQ AQM) for handling LS and non-LS traffic. In embodiments, LS capable deviceis configured to transmit congestion signals to wireless network. For example, a congestion signal may include an ECN signal, such as ECN-capable packets. It should be noted that wireless networkis described as receiving the congestion signal from a single LS capable devicefor ease of description, and as such, the congestion signal may include signals received from a plurality of LS capable devices.

202 4 223 4 223 4 202 4 223 4 223 In embodiments, wireless networkincludes non-LS capable device. Non-LS capable devicemay include any device not capable of supporting LS. In instances, wireless networkmay include a plurality of non-LS capable devices. In embodiments, non-LS capable deviceis connected to wireless network through a default connection. In some embodiments, the default connection may include a connection using a network slice.

202 4 222 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Wireless network, based on detecting a congestion signal from LS capable device, makes adjustments to network configurations including non-LS capable access node and devices. The adjustments may include prioritizing low-bitrate slices for non-LS capable access node(s); adjusting power control settings for UL for non-LS capable access node(s); adapting slice configurations, such as slice type, slice throughput, radio spectrum, and isolation level, for the non-LS capable access node(s); prioritizing essential (e.g., first responder, disaster relief, etc.) non-LS traffic; reducing UL layers for non-LS capable access node(s); assigning non-LS capable wireless device(s) to non-LS capable access node(s); moving (handover) the non-LS capable wireless device(s) from the LS capable access node(s) to non-LS capable access node(s); adjusting UL and download (DL) resource allocation to optimize for stability for example, by using one or more KPIs for an access node, such as call set up success rate, dropped call rate, site data throughput, of the non-LS capable access node(s); and adjusting cell boundaries of LS and non-LS capable access node(s). Furthermore, the adjustments may be any combination of the adjustments listed above.

202 260 260 260 4 4 4 4 4 223 In embodiments, wireless networkmay exchange congestion metrics with a datastore. Datastoremay include any data storing medium, such as a database. In instances, data related to detected congestion signals may be stored in datastoreto be used as training data for machine learning processes. For example, data related to marked packets (i.e. congestion signal) may be correlated to timestamps for the signals, latency, ECN marking levels, and packet loss from LS capable access nodes and LS capable wireless devices. In embodiments, this training data may be used for training machine learning models, which may be used for predicting congestion in the network prior to or after detection of LS congestion signals. For example, if network congestion is predicted adjustments or combinations of adjustments described above may be made to non-LS capable access nodes. In an example, training data may include data related to frequency and duration of congestion events, such as congestion signals correlated to the length of time of detection of the signals. Training data may include network traffic volume during detection of signals. In an example, training data may also include packet loss for non-LS device.

3 FIG. 1 FIG. 300 300 305 4 4 4 4 4 4 171 4 172 With reference to, a flow diagram of methodfor network congestion management is presented. Methodincludes, at step, receiving, by a wireless network communicatively connected to an LS capable access node and a non-LS capable access node, a congestion signal from an LS capable wireless device. In embodiments, the congestion signal may include an ECN. The LS capable access node and non-LS capable access node may include, respectively, LS capable access nodeand non-LS capable access nodedescribed in reference to.

310 300 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 At step, methodincludes adjusting, by the wireless network, network configurations for a non-LS capable access node based on the congestion signal. The adjustment to network configurations for a non-LS capable access node may include any combination of the following: prioritizing low-bitrate slices for non-LS capable access node(s); adjusting power control settings for UL for non-LS capable access node(s); adapting slice configurations for the non-LS capable access node(s); prioritizing essential non-LS traffic; reducing UL layers for non-LS capable access node(s); assigning non-LS capable wireless device(s) to non-LS capable access node(s); moving (handover) the non-LS capable wireless device(s) from the LS capable access node(s) to non-LS capable access node(s); adjusting UL and download (DL) resource allocation to optimize for stability of the non-LS capable access node(s); and adjusting cell boundaries of LS and non-LS capable access node(s).

3 FIG. 350 350 315 4 4 4 4 With reference to, a flow diagram of methodfor network congestion management using machine learning is presented. Methodincludes, at step, collecting historical congestion data from LS capable devices and access nodes. The historical congestion data may be collected from LS devices and access nodes throughout the network. The historical congestion data may include any combination of data related to ECN marked packets (i.e. congestion signal) correlated to timestamps for the signals, latency, ECN marking levels, and packet loss, frequency, duration of congestion events and time length congestion signals from LS capable access nodes and LS capable wireless devices.

320 4 223 At step, a machine learning model is trained using training data comprising collected historical congestion data. In an example, training data may include data related to frequency and duration of congestion events, such as congestion signals correlated to the length of time of detection of the signals. Training data may include network traffic volume during detection of signals. In an example, training data may also include packet loss for non-LS device.

325 4 4 4 4 4 4 171 4 172 1 FIG. At step, the method includes receiving, by a wireless network communicatively connected to an LS capable access node and a non-LS capable access node, a congestion signal from an LS capable wireless device. In embodiments, the congestion signal may include an ECN. The LS capable access node and non-LS capable access node may include, respectively, LS capable access nodeand non-LS capable access nodedescribed in reference to.

330 325 320 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 At step, the method includes inputting the congestion signal received in stepinto the model trained in stepto adjust by the wireless network, network configurations for a non-LS capable access node. The adjustment to network configurations for a non-LS capable access node may include any combination of the following: prioritizing low-bitrate slices for non-LS capable access node(s); adjusting power control settings for UL for non-LS capable access node(s); adapting slice configurations for the non-LS capable access node(s); prioritizing essential non-LS traffic; reducing UL layers for non-LS capable access node(s); assigning non-LS capable wireless device(s) to non-LS capable access node(s); moving (handover) the non-LS capable wireless device(s) from the LS capable access node(s) to non-LS capable access node(s); adjusting UL and download (DL) resource allocation to optimize for stability of the non-LS capable access node(s); and adjusting cell boundaries of LS and non-LS capable access node(s)

4 FIG. 1 FIG. 400 400 400 491 492 491 492 491 Now referring to, an example computing deviceis presented. In embodiments, computing devicemay include a node device, such as devices operating within communication network described in reference to. In this example, computing deviceincludes at least one processorcommunicably coupled to a computer-readable storage medium. The at least one processormay include a microprocessor, a microcontroller, one or more central processing unit (CPU) cores, an application-specific integrated circuit (ASIC), one or more graphical processing unit (GPU) cores, a field programmable gate array (FPGA), and/or any other hardware device suitable for retrieval and execution of instructions from computer-readable storage medium. In instances, at least one processormay include electronic circuitry for performing instructions described in this disclosure.

492 492 492 400 492 3 FIG. In instances, computer-readable storage mediummay be any medium suitable for storing executable instructions. In examples, without limitation, computer-readable storage mediummay include read-only memory (ROM), random-access memory (RAM), erasable electrically programmable ROM (EEPROM), Solid State Drive (SSD), optical disc, and the like. Computer-readable medium storagemay be disposed within computing device. In embodiments, computer-readable storage mediummay be external, and communicably connected, to a computing device. The instruction stored on computer-readable storage medium may be used to implement method steps described in reference to.

492 493 494 In this example, computer-readable storage mediumis encoded with a set of instructionsand. In embodiments, executable instructions included in each block may be included in different blocks shown and blocks not shown.

493 491 491 Instruction, when executed by at least one processor, configures the at least one processorto receive congestion signal(s) s.

494 491 491 Instruction, when executed by at least one processor, configures the at least one processorto adjust network configuration for the non-L4S capable access node.

392 391 4 5 FIG. In embodiments, computer-readable storage mediummay include instructions configuring the at least one processorto determine the session configuration used for adjusting the network configurations for the non-LS capable access node as a function of a machine leaning model. The machine learning model is described in further detail in reference to.

5 FIG. 500 502 504 506 502 504 502 504 Now referring to, an example processing node, which may be configured to perform the methods and operations disclosed herein for network congestion management. The processing node 500 includes a communication interface, user interface, and processing systemin communication with communication interfaceand user interface. Communication interfacemay include hardware components, such as network communication ports, devices, routers, wires, antenna, transceivers, etc. User interfacemay include hardware components, such as touch screens, buttons, displays, speakers, etc.

506 508 510 510 510 512 500 512 508 512 510 506 500 502 500 504 500 500 512 4 FIG. Processing systemincludes a central processing unit (CPU) or processorand storage. Storagemay include a disk drive, flash drive, memory circuitry, or other memory device including, for example, a buffer. Storagecan store softwarewhich is used in the operation of the processing node. Softwaremay include computer programs, firmware, or some other form of machine-readable instructions, including an operating system, utilities, drivers, network interfaces, applications, or some other type of software. Processing system 506 may include a processorand other circuitry to retrieve and execute softwarefrom storage, which may be internal or external to the processing system. Processing nodemay further include other components such as a power management unit, a control interface unit, etc., which are omitted for clarity. Communication interfacepermits processing nodeto communicate with other network elements. User interfacepermits the configuration and control of the operation of processing node. Processing nodemay be included in various elements of the wireless network including an access node, proxy call session control function (P-CSCF), gateway mobile location center (GMLC), radio resource control (RRC), inter-cell interference coordination (ICIC), medium access control (MAC), session border controller (SBC), and the like. In this example, softwaremay include the instructions described in reference to.

512 513 513 513 500 4 4 172 513 1 FIG. In embodiments, softwareincludes machine learning processes. In embodiments, processing system 506 may use machine learning processesto perform determinations, classifications and or analysis steps. In embodiments, machine learning processesmay be used to generate a machine learning model. For example, processing nodemay be configured used for adjusting network configurations for the non-LS capable access nodes, such as non-LS capable access nodedescribed in reference to, using machine learning processes.

506 513 506 513 513 In instances, processing systemmay use machine learning processesto generate training data. In embodiments, processing systemmay use the training data to train a machine learning model. For example, machine learning processesmay model relationships between two or more categories of data elements using the training data. In embodiments, training data may include historical congestion data from L4S capable access nodes and wireless devices. In some examples, training data may include one or more elements not categorized. In embodiments, machine learning processesmay include a neural network. As used herein, a neural network is a network of data structures, or nodes, that contains one or more inputs, one or more outputs and a function for determining outputs based on the inputs, where the network includes an input layer, an output layer and oner or more intermediate layers. In instances, neural network may have a recurrent architecture, such as a recurrent neural network (RNN). For example, RNN may be used for generating time series predictions. In embodiments, the neural network may have a memory cell based architecture, such as a long short-term memory (LSTM) network. For example, LSTM may be used for generating time series forecasting based on long term dependencies.

Although the descriptions provided herein may be in the context of certain radio access technologies, networks, and network topologies, such as 5G/NR mobile communications, the proposed concepts, schemes, and any variations thereof may be implemented in, for and by other types of radio access technologies, networks, and network topologies. Such radio access technologies, networks, and network topologies may include, for example and without limitation, Long-Term Evolution (LTE), Internet-of-Things (IoT), Narrow Band Internet of Things (NB-IoT), vehicle-to-everything (V2X), fixed wireless internet, and non-terrestrial network (NTN) communications. Thus, the scope of the disclosure is not limited to the examples described herein.

The exemplary systems and methods described herein may be performed under the control of a processing system executing computer-readable codes embodied on a computer-readable recording medium or communication signals transmitted through a transitory medium. The computer-readable recording medium may be any data storage device that can store data readable by a processing system, and may include both volatile and nonvolatile media, removable and non-removable media, and media readable by a database, a computer, and various other network devices. Examples of the computer-readable recording medium include, but are not limited to, read-only memory (ROM), random-access memory (RAM), erasable electrically programmable ROM (EEPROM), flash memory or other memory technology, holographic media or other optical disc storage, magnetic storage including magnetic tape and magnetic disk, and solid-state storage devices. The computer-readable recording medium may also be distributed over network-coupled computer systems so that the computer-readable code is stored and executed in a distributed fashion. The communication signals transmitted through a transitory medium may include, for example, modulated signals transmitted through wired or wireless transmission paths.

The above description and associated figures teach the best mode of the invention. The following claims specify the scope of the invention. Note that some aspects of the best mode may not all be within the scope of the invention as specified by the claims. Those skilled in the art will appreciate that the features described above can be combined in various ways to form multiple variations of the invention. As a result, the invention is not limited to the specific embodiments described above, but only by the following claims and their equivalents.